Heritability, Correlation and Path Coefficient Analysis in Twenty Ber Genotypes

Transcription

1 Academic Journal of Plant Sciences 3 (): 9-98, 010 ISSN IDOSI Publications, 010 Heritability, Correlation and Path Coefficient Analysis in Twenty Ber Genotypes M.N. Islam, M.M. Hossain, M.M. Rahman, M.S. Uddin and M.M. Rohman 1 Horticultural Research Centre, BARI, Gazipur 1701, Bangladesh Department of Horticulture, BSMRAU, Gazipur 1706, Bangladesh 3 Plant Breeding Division, BARI, Gazipur 1701, Bangladesh Abstract: A study was undertaken to analyses the heritability, correlation and path co-efficient for growth and fruit characters in twenty ber (Ziziphus mauritiana Lam) genotypes grown at Fruit Research Farm under Horticulture Research Centre, BARI during January to June 006. The highest range of variation was recorded in fruit weight ( g), followed by yield per plant ( kg), stone weight ( g) and pulpstone ratio ( ). The highest GCV (55.67 %) and PCV (55.67 %) were recorded for fruit weight while both were lowest for stone breadth (16.4 % and 17.1 %). High heritability coupled with high or moderate degree of genetic advance was estimated in fruit weight (99.35 % and ), yield per plant (95.38 % and 93.38), stone weight (99.05 % and 90.8), pulp/stone ratio (98.74% and 66.34), fruit breadth (90.14 % and 49.70), stone length (98.34% and 51.83) and number of fruits per plant (89.90 % and 49.63). Strong positive correlation was found between yield per plant and individual fruit weight (r g=0.965), stone weight (r g=0.74), fruit length (r g=0.737), fruit breadth (r g=0.807) and pulp-stone ratio (r g=0.574). Among the selected characters, stone weight had the highest positive direct effect (1.777) on fruit yield followed by pulp/stone ratio (1.455), fruit breadth (0.35) and fruit length (0.107). Key words: Ziziphus mauritiana Lam Correlation heritability Path coefficient INTRODUCTION be useful in selecting plants with desirable characters to develop new varieties. But information on these aspects Ber (Ziziphus mauritiana Lam) is an amazing fruit in ber is meager in the country. Therefore, there is a need crop in Bangladesh. An increasing trend of it s to estimate the genetic variability in ber through production is quite visible in the country. In , the biometrical procedures which may be useful to develop production of ber was tons as against tons some selection criteria in the improvement program of ber. [1]. One of the important reasons of this increasing trend The knowledge of association of plant characters as is the development of good quality varieties, especially determined by the correlation coefficient is helpful for BU kul-1 (Apple kul) and BAU kul-1. There is still selection of desirable characters under a breeding necessity to develop more varieties of this potential fruit. program. Thus measurements of correlation coefficient More varieties may be developed exploiting the wide between characters are a matter of considerable genetic variability exists in ber through breeding program. importance in selection indices and also permit the The progress of breeding is conditioned by the prediction of correlated response []. Some workers in magnitude, nature and interaction of genotypic and other countries especially in India have reported extensive environmental variations in the plant characters. It then work on correlation and path coefficients [3-5]. However, becomes necessary to partition the observed variability very little attempts have been taken, specifically on fruit into its heritable and non-heritable components with the crops in Bangladesh. help of suitable genetic parameters such as genetic The estimate of path co-efficient analysis is important coefficient variation, heritability estimates and genetic for better understanding of the crop. It gives specific advance etc. This will provide valuable information on the measure on the direct and indirect effect of each mode of inheritance of different characters, which would component character upon fruit yield. The present study Corresponding Author: M.N. Islam, Horticultural Research Centre, BARI, Gazipur 1701, Bangladesh. 9

2 Acad. J. Plant Sci., 3 (): 9-98, 010 was, therefore, undertaken to estimate the extent of genotypic and phenotypic variability, heritability and genetic advance, correlation coefficient among the selected characters and direct and indirect effects of component characters on yield of ber. MATERIALS AND METHODS Materials: Twenty diverge genotypes of ber (Ziziphus mauritiana Lam) grown at Fruit Research Farm under Horticulture Research Centre, BARI were selected for conducting this study. The plants were previously established by in situ budding through top-working. The ages of the stock of the plants were eight to ten years old. The experiment was premeditated in Randomized Complete Block Design (RCBD) with three replications. A single tree of each cultivar/variety having uniform size and approximately same age (8-10 years old) considered as unit of replication. Methods: Data on twelve quantitative characters of 0 ber genotypes were used for variability estimation, correlation and path coefficient analysis. The selected characters were plant height, canopy spread in North-South direction, canopy spread in East-West direction, fruit weight, fruit length, fruit breath, stone weight, stone length, stone breath, pulp/stone ratio, number of fruits per tree and yield per tree. Statistical Analysis: The analysis of variance for each character was performed using Microsoft Excel and MSTAT software. Estimation of Variation and its Heritable Components in Genetic Parameters: Phenotypic and Genotypic Variance: The phenotypic and genotypic variances were estimated according to Johnson et al. (1955). The error mean square (EMS) was considered as error variance ( e). Genotypic variances ( g) were calculated by GMS-EMS/r, where, GMS= Genotypic mean square, EMS= Error mean square r=replication. The phenotypic variances ( p) were derived by adding genotypic variances with the error variances ( e) as given by the following formula: p = g+ e. Genotypic and Phenotypic Co-efficient of Variation (GCV and PCV): Genotypic and phenotypic co-efficient of variations were computed by the formula suggested by Burton and Devane [6] as given below: g Genotypic co-efficient of variation (GCV) = 100 X Where, g = Phenotypic standard deviation, X = population mean.0 P Similarly, The phenotypic co-efficient (PCV = 100 X Heritability: Heritability in broad sense was estimated according to Johnson et al. [7]. g %H = 100 P Genetic Advance: The expected genetic advance for different characters was estimated as per the formula given by [7]. g Genetic advance = Kx 100 P Where, K = Selection differential, the value of which is.06 at 5% selection intensity, p = Phenotypic standard deviation. Character Association by Correlation Study and Path Analysis: Correlation and path-coefficient analysis were estimated by the association of characters and causeeffect relationship studied for yield and component characters. Estimation of Correlation: Association of different characters under the study was analyzed by the working out genotypic and phenotypic degree of correlation and simple correlation coefficient for all the possible parts of character combination by the method of Hayes et al. [8] and Al-Jabouri et al. [9]. Estimation of Direct and Indirect Effect of Different Characters on Yield: In order to find a clear picture of the inter-relationship between fruit yield and other components, path analysis splits the correlation coefficient into the measure of the direct and indirect effect of each contributing characters towards yield at genotypic level was done following [10]. Calculation of Residual Effect: After calculating the direct and indirect effect of different characters, the residual effect was calculated using the formula suggested by Singh and Choudhury [11]. 93

3 Acad. J. Plant Sci., 3 (): 9-98, 010 RESULTS AND DISCUSSION Genotypic and Phenotypic Variances: The genotypic and phenotypic variance ranged from and Estimation of Variation and its Heritable Components in to and 0.8 (Table 1). Estimates of Genetic Parameters: The extent of variability in respect genotypic and phenotypic variances were high for to twelve characters in different genotypes, measured in number of fruits harvested per plant, individual terms of range, genotypic coefficient of variation (GCV), fruit weight and yield per plant. Genotypic and phenotypic coefficient of variation (PCV) along with the phenotypic component of variance were the highest heritability, expected genetic advance are given in for number of fruits per plant followed by yield per Table 1. plant ( and ). It was and for individual fruit weight. The moderate component of Range and Mean: The highest range of variation was variance was found for pulp-stone ratio (10.65 and 10.78). recorded in fruit weight ( g), followed by yield The phenotypic component of variance was higher per plant ( kg), stone weight ( g) and than the genotypic component of variance for pulp-stone ratio ( ) with the mean of 18.3 g, number of fruits per plant ( and ) 8.77 kg, 1.63 g and 10.07, respectively among the and yield per plant ( and ). Compare to characters (Table 1). Moderate range of variation was genotypic component of variance, a little higher found in number of fruits per plant ( ) with the phenotypic component of variances were also found in mean of The remaining contributing characters fruit weight, fruit length, fruit breadth, stone weight, had narrow range of variation indicating narrow range of stone length, stone breadth and pulp-stone ratio variability among the ber genotypes for these traits. (Table 1). The differences of phenotypic and genotypic Nanohar et al. [1] observed a wide range of variances were found low for the characters fruit weight, variability in fruit weight ( g), pulp-stone ratio (5.5- fruit length, fruit breadth, stone weight, stone length, 14.0) and yield per plant ( kg) among 13 ber stone breadth and pulp-stone ratio indicating that these genotypes in their study. Saran et al. [6] also observed characters were less influenced by environment. The the highest range of variation for fruit yield per plant and phenotypic and genotypic variances for the character fruit weight. Vijay and Manohar [13] stated that characters number of fruits per plant were and which showed high range of variation should be given respectively and the difference was slight high priority in the selection. Since in breeding program, suggesting that the character was influenced by variability among the population is a pre-requisite, high environment. In respect to error variance, genotypic variability observed in respect of the traits under study variances were high in all the characters except number of implies that there is scope for making effective fruits further indicating that the later was influenced by improvement of these traits [14]. environment (Table 1). Table 1: Estimates of genetic parameters for different characteristics in ber genotypes Variance components Range Genotypic Phenotypic Eroor GCV PCV Herita- Genetic Character Min. Max. Mean variance variance variance (%) (%) bility(%) advance Plant Height Canopy spread (N-S) Canopy spread (E-W) Fruit Weight Fruit length Fruit Breadth Stone Weight Stone Length Stone Breadth Pulp-stone ratio Number of fruits/ plant Yield/plant

4 Acad. J. Plant Sci., 3 (): 9-98, 010 Comparatively higher phenotypic component of genetic advance for yield per plant (91.6% and 61.16) and variance than the genotypic component and the extent of pulp/stone ratio (91.67 % and 56.91) in ber. latter component showed that quantitative characters of With high value of heritability coupled with moderate ber fruits are mostly hereditary in nature, which is evident degree of genetic advance were recorded for fruit breadth by higher values recorded [4]. Nanohar et al. [1] reported (90.14 % and 49.70), stone length (98.34% and 51.83) and that genotypic component of variances were greater than number of fruits per plant (89.90 % and 49.63) (Table 1). the environmental components for fruit breadth, T.S.S and High heritability value along with high or moderate value pulp/stone ratio whereas it was lesser in case of fruit of genetic advance would be most effective condition for length and fruit weight in ber which indicated fruit length selection. Such condition arises due to action of additive and fruit weight were influenced by environmental genes [15]. variations as compared to other fruit characters. Moderate heritability coupled with moderate degree of genetic advance was recorded for fruit length (54.1 % Genotypic and Phenotypic Co-efficient of Variation: and 35.56). However, Low heritability along with low The genotypic and phenotypic co-efficient of variations genetic advance as exhibited due to polygenic inheritance (GCV and PCV) are the measures of variability among the was not found under study. genotypes under study. The genotypic co-efficient of The perusal of the result revealed that as higher variation (GCV) measuring the range of genetic variability estimates of fruit weight, stone weight, pulp/stone for different plant characters helps to compare this ratio, yield per plant as well as stone length, fruit variability and phenotypic co-efficient of variation (PCV) breadth, number of fruits per plant, with respect to indicated the interaction effect of environment on these GCV, heritability and genetic advance indicated additive traits. For fruit characters, the highest GCV and PCV were gene effects controlling these traits, individual plant found for fruit weight (55.67 and 55.85) followed by yield selection for these traits would be effective in ber. per plant (46.47 and 47.59), stone weight (44.30 and 44.51) and pulp/stone ratio (3.41 and 3.61) while stone breadth (16.4 and 17.1) was minimum. Less difference between Character Association by Correlation and Path Co- GCV and PCV recorded for all the fruit characters efficient Analysis: For a sound-breeding programme, except fruit length further revealed that these characters information on the genetic association between yield and were least influenced by environment, as well it indicated its components is a pre-requisite. From this point of view, low variability of these characters within the genotypes. the relationship between yield of ber and 11 another The GCV and PCV for fruit length were 3.45 and important characters were endeavored to find out through which indicated variability remained within the genotypes, correlation and path co-efficient analysis. however, not much influenced by the environment. The high GCV can be exploited by appropriate CORRELATION selection. Singh and Jalikop [14] reported that higher the value of genotypic co-efficient of variation more amenable Correlation study among different yield contributing the character for improvement. Nanohar et al. [1] found characters revealed that all the genotypic correlation moderate value of GCV for yield per plant (31.01) in ber. coefficients were higher than the phenotypic correlation co-efficient (Table ). Heritability and Genetic Advance: From Table 1. high Highly significant positive genotypic and phenotypic heritability (per cent of mean) estimates coupled with high correlation was observed for plant height with canopy genetic advance in fruit weight (99.35 % and ), yield spread both in North-South (0.976 and 0.748) and Eastper plant (95.38 % and 93.38), stone weight (99.05 % and West direction (0.946 and 0.718). Positive and highly 90.8) and pulp/stone ratio (98.74% and 66.34) which significant genotypic correlation was found between plant indicate that these characters were less influenced by height and number of fruits per plant (0.69). Plant height environment demonstrating either these were simply also showed positive and significant genotypic inherited characters governed by a few major genes or correlation with yield per plant (0.501). However, this additive gene effect even if, they were under polygenic correlation was positive but insignificant in phenotypic control and therefore, selection of these characters aspect (0.40). These results indicated that canopy area, would be more effective for yield improvement [7, 15]. number of fruits and yield per plant would be increased Nanohar et al. [1] found high value of heritability and with increasing of plant height. 95

5 Acad. J. Plant Sci., 3 (): 9-98, 010 Table : Genotypic (above diagonal) and phenotypic (below diagonal) correlation coefficient among various characters studied in 0 ber genotypes Plant Canopy Canopy Fruit Fruit Fruit Stone Stone Stone Pulp/ Yield/ height (N-S) (E-W) weight length breadth weight length breadth Stone Fruits / plant Character (cm) (cm) (cm) (g) (cm) (cm) (g) (cm) (cm) ratio plant (kg) Plant height (cm) ** ** ** * Canopy (N-S) (cm) ** 0.78 ** ** Canopy (E-W) (cm) ** ** ** * Fruit weight (g) ** ** ** ** ** 0.67 ** ** ** Fruit length (cm) ** ** ** ** * ** ** ** Fruit breadth (cm) ** ** ** ** * ** ** ** Stone weight (g) ** * ** ** ** * 0.74 ** Stone length (cm) ** ** ** ** * ** ** ** Stone breadth (cm) ** * ** ** * Pulp/Stone ratio ** * ** ** * ** Fruits /plant * * 0.53 * ** * * ** * * 0.85 Yield/plant (kg) ** ** ** 0.78 ** ** * * 0.05 * = Indicate 5 % level of significance ** = Indicate 1 % level of significance Positive correlation was observed for canopy spread significant correlation in genotypic level (0.500) and with number of fruits per plant (0.690 and for N-S positive but insignificant correlation at phenotypic aspect direction and and for E-W direction.). Canopy (0.389). These results indicate that, there was a positive spreads had positive correlation with yield per plant, even association among the fruit and seed characters coupled this relation was found to be significant in genotypic level with yield per plant. (0.506) by canopy spread in East-West direction Highly significant positive genotypic and phenotypic (Table ). However, this was insignificant both in correlation was observed for stone weight with fruit genotypic (0.381) and phenotypic (0.85) components by weight (0.800 and 0.800), fruit breadth (0.637 and 0.610), canopy spread in North -South direction. The assorted stone breadth (0.788 and 0.765) and yield per plant (0.74 results might be indicating the plausible involvement of and 0.78). With fruit length, this correlation was found to microclimatic factors. be also significant, but with pulp-stone ratio it was Single fruit weight had positive and highly significant insignificant in both levels (0.109 and 0.106). This correlation with fruit length (0.851 and 0.641), fruit breadth indicates that higher edible portion of ber was not much (0.911 and 0.874), stone weight (0.800 and 0.800), stone dependent on stone weight. length (0.849 and 0.848), stone breadth (0.607 and 0.594), Highly significant correlation was found for stone pulp/stone ratio (0.67 and 0.669) and yield per plant length with pulp/stone ratio (0.645 and 0.638) and yield (0.905 and 0.890). This result suggested that increase in per plant (0.698 and 0.691). Stone breadth had positive but the fruit weight increases the fruit yield. It also indicates insignificant correlation with pulp/stone ratio (0.01 and that, there was a strong and positive association with fruit 0.08). It has positive and significant correlation with yield weight and seed weight. These were in line with the per plant (0.544 and 0.535). findings of Saran et al. (007). They observed that yield Highly significant correlation was found between of ber genotypes had highly significant and positive pulp/stone ratio and yield per plant in genotypic correlation with fruit weight. component (0.574). In case of phenotypic component, this Positive and significant correlation was observed correlation was found to be positive and significant for fruit length with fruit breadth (0.880 and 0.648), (0.560). stone length (0.860 and 0.658), pulp/stone ratio Positive correlation was recorded for number of fruits (0.677 and 0.495) and yield (0.737 and 0.568) while it was per plant with yield per plant (0.85 and 0.05), but this negative and significant with number of fruits per correlation was not found to be significant. With fruit and plant (-0.69) in genotypic level and negative but seed characters, number of fruits per plant had negative insignificant in phenotypic expression (-0.391). With correlation, even most of the cases these correlations stone breadth, fruit length showed positive and were found to be significant (Table ). 96

6 Acad. J. Plant Sci., 3 (): 9-98, 010 Table 3: Direct and indirect effects (bold) of component characters on yield of ber genotypes Plant Canopy Canopy Fruit Fruit Fruit Stone Stone Stone Pulp/ height (N-S) (E-W) weight length breadth weight length breadth Stone Fruits / Character (cm) (cm) (cm) (g) (cm) (cm) (g) (cm) (cm) ratio tree Plant height (cm) Canopy (N-S) (cm) Canopy (E-W) (cm) Fruit weight (g) Fruit length (cm) Fruit breadth (cm) Stone weight (g) Stone length (cm) Stone breadth (cm) Pulp/Stone ratio Fruits /tree Residual value = 0.48 Path Coefficient Analysis: In order to find a clear picture From the result of the path analysis in Table 3. it of the inter-relationship between fruit yield and other revealed that stone weight had the highest positive direct components path coefficient analysis has been performed effect (1.777) on fruit yield followed by pulp/stone ratio where yield of ber was considered as resultant variable (1.455), fruit breadth (0.35) and fruit length (0.107). Saran and the rest characters as causal variable. et al. (007) reported that fruit weight had the highest Among the different characters plant height showed direct and positive effect (0.9987) on yield per tree direct negative effect (-0.684) on yield per plant which followed by fruit size (0.3375) and stone size (0.567) in was compensated by indirect and significant positive ber. (0.343) effect through number of fruits per plant via In some cases, the path coefficient analysis gave a indirect positive effect (0.660) of canopy spread in North- somewhat different picture than the simple correlation south direction (Table 3). Fruit weight exhibited highly analysis. The apparent contradiction is due to the fact significant association with yield (r g=0.905). Whereas, it that the correlation simply measures mutual association had negative direct effect (-0.979) on fruit yield possibly without regard to causation, whereas the path coefficient via indirect negative effect of number of fruits per plant analysis specifies the causes and measures their relative (-0.347) which was compensated by positive indirect importance. For example, correlation coefficient between effects of stone weight (1.4) as well as fruit length single fruit weight and yield per plant (r g=0.905) gave an (0.091) and fruit breadth (0.31) that contribute to higher impression that fruit weight had positive influence on fruit edible portion as evident by direct positive effect of weight. The path coefficient analysis, however, revealed pulp-stone ratio (0.978). A low positive direct effect that fruit weight had negative direct effect on fruit yield (0.107) was exhibited by fruit length on yield per plant, though it compensated through positive effects by stone mostly of which contributed through positive effect weight and fruit size. (0.986) of increasing pulp/stone ratio with increasing of fruit length. Fruit breadth had positive direct effect (0.979) CONCLUSION on fruit yield via positive indirect effect (1.093) of pulpstone ratio. Highly positive direct effect (1.777) on fruit Considerable genetic variability was present among yield was exhibited by stone weight. Stone length showed the ber genotypes selected in the study. The highest negative direct effect (-0.579) on yield. Stone breadth had range of variation was recorded in fruit weight, followed low negative direct effect (-0.031) on yield. Harvested by yield per plant, stone weight and pulp-stone ratio. fruits per plant showed positive but insignificant High heritability estimates coupled with high or moderate correlation with yield (r g=0.85) which might be due to degree of genetic advance was estimated in fruit weight, positive indirect effect of this trait via fruit weight (0.48). stone weight, pulp/stone ratio, fruit breadth, stone length However, most of the yield contributing characters exhibit and number of fruits per plant. Both correlation and path negative result at interaction with number of harvested co-efficient analyses carried in this study suggested that fruits per plant. fruit weight, stone weight and number of fruits per plant 97

7 Acad. J. Plant Sci., 3 (): 9-98, 010 are major components of fruit yield. The above all 7. Johnson, H.W., H.F. Robinson and R.F. Comstock, estimation of heritability, correlation and path-coefficient Estimates of genetic and environmental analysis of ber genotypes indicated that individual plant variability in soybeans. Agron. J., 47: selection would be effective for varietal improvement of 8. Hayes, H.K., F.R. Immer and D.C. Smith, this crop. nd Methods of Plant Breeding. ( ed.). McGraw Hill REFERENCES 9. Book Co. Inc. New York. pp: 55I. Al-Jabouri, R.A., P.A. Miller and H.F. Robinson, Genotypic and environmental variance in 1. BBS, 007. Year Book of Agricultural Statistics of upland cotton cross of interspecific origin. Bangladesh. Bangladesh. Bureau of Statistics, Agron J., 50: Ministry of Planning, Govt. of the Peoples Republic 10. Dewey, D.R. and K.H. Lu, A correlation and of Bangladesh, Dhaka. pp: 93. path co-efficient analysis of components of crested. Lerner, M., The genetic basis of selection. John wheat grass seed production. Agron. J., 51: Willey and sons. New York. pp: Singh, R.K. and B.D. Choudhury, Biometrical 3. Sidhu, A.S. and J.S. Brar, Correlation and path methods in quantitative genetic analysis. Kalyani coefficient analysis for yield, quality and earliness in Publishers, New Denhi, pp: watermelon (Citrllus lantus Thunb. Mansf.). Indian 1. Nanohar, M.S., N.L. Sen and J.P. Yadvendra J. Agric. Res., 15(1): Phenotypic variation and its heritable components in 4. Attri, B.L., T.V.R.S. Sharma, D.B. Singh and P. some biometric characters in ber (Zizyphus Nagesh, Genetic variability and correlation mauritiana Lamk.). Indian J. Hort., 47(l): studies in mango collections of South Andaman. 13. Vijay, O.P. and M.S. Manohar, Studies on Indian J. Hort., 56(): genetic variability, correlation and path analysis in 5. Saran, P.L., A.K. Godara, G. Lal and I.S. Yadav, 007. okra (Abelmoschus esculentus (L.) Moench). Indian Correlation and path coefficient analysis in ber J. Hort., 47(1): genotypes for yield and yield contributing traits. 14. Singh, R. and S.H. Jalikop, Studies on Indian J. Hort., 64(4): variability in grape. Indian J. Hort., 47(1): Burton, G.W. and E.H. de Vane, Estimating the 15. Panse, V.G., Genetics of quantitative characters heritability in tall fescue (Festuca arundinaced) from in relation to plant breeding. Indian J. Genet. Plant replicated clonal material. Agron. J., 45: Breed., 17: